Retrospectives: The Introduction of the Cobb–Douglas Regression

At the 1927 meetings of the American Economic Association, Paul Douglas presented a paper entitled “A Theory of Production,” which he had coauthored with Charles Cobb. The paper proposed the now familiar Cobb–Douglas function as a mathematical representation of the relationship between capital, labor, and output. The paper's innovation, however, was not the function itself, which had originally been proposed by Knut Wicksell, but the use of the function as the basis of a statistical procedure for estimating the relationship between inputs and output. The paper's least squares regression of the log of the output-to-capital ratio in manufacturing on the log of the labor-to-capital ratio—the first Cobb–Douglas regression—was a realization of Douglas's innovative vision that a stable relationship between empirical measures of inputs and outputs could be discovered through statistical analysis, and that this stable relationship could cast light on important questions of economic theory and policy. This essay provides an account of the introduction of the Cobb–Douglas regression: its roots in Douglas's own work and in trends in economics in the 1920s, its initial application to time series data in the 1927 paper and Douglas's 1934 book The Theory of Wages, and the early reactions of economists to this new empirical tool.